Inus Dreckmeyr

While the hype around the Internet of Things (IoT) and cloud computing is enough to put most sane business people off the concept, Netshield CEO Inus Dreckmeyr advises against dismissing IoT as just another new-fangled way for vendors to make money.

“It is easy to fall into one of two traps with IoT: either the hype turns business management teams off IoT completely, or it encourages them to recklessly add sensors and IP connections to everything. Both approaches are a good way to fail to extract the business value inherent to the technologies that are deployed in IoT implementations,” Dreckmeyr says.

He suggests that companies keep in mind the following when investigating their own IoT implementations:

  1. Don’t confuse “cloud” with “IoT”

The cloud is a network of servers, and each server has a different function. Like IoT it is built on IP technology, but is separate from the Internet. The IoT, on the other hand, is a network of sensors and control mechanisms that can live anywhere and have been around for ages. Their ubiquity has simply made these sensors and control devices more affordable, and has meant they are now included in objects that wouldn’t traditionally include sensors.

  1. Define what IoT means to your company

There are as many definitions of IoT as there are commentators and research houses. Some definitions include every single consumer device that pings a gaming server, others limit it to the use of objects in industry, and others basically confine it to what used to be known as telemetry. (This, Dreckmeyr says, also explains why projections of the IoT’s size vary so dramatically. Predictions for the market size in 2020 vary from the IDC’s 2015 projection of 29.5 billion “IoT end-points” to Gartner’s 2013 prediction of 20.8 billion “connected things” to ABI Research’s 2014 projection of 41 billion “active wireless connected devices”!)

“Decide which definition works best for your company. Then apply IoT principles to extend an ongoing system or process, rather than basing a newly-developed application on IoT only.”

  1. Decide what should – and shouldn’t – connect to the Internet

It is tempting to think that all sensors should be Internet-capable, but it is often good enough for a sensor to be connected to a local hub, which processes the data it receives from the sensor, and passes it along for action. “For instance, in a data centre, it is completely acceptable for a temperature gauge to connect to a local command centre that then sends instructions to an air-conditioner to activate, while alerting the IT team via SMS or email that there was a concern about temperature controls in the datacentre. The sensor itself doesn’t have to be IP-capable, only the hub”, Dreckmeyr explains.

  1. Don’t assume you need to employ a data scientist – but don’t skimp on analysis either

“There is absolutely no point in employing a data scientist to analyse temperature data from a data centre in isolation,” Dreckmeyr explains. “But if you want to leverage the temperature data you have against other data sets, do employ a professional – either permanently or outsourced – to do the analysis for you.”

  1. Don’t assume that you can always replace humans

This correlates to the above point. Says Dreckmeyr: “There comes a point where human intervention will be required, either because the decision based on the data from IoT devices isn’t straightforward, or because it is a matter of such importance that it cannot be left to a machine that doesn’t understand ethics or human emotions. Even if that machine has built-in artificial intelligence, advanced machine learning, a digital twin and intelligent apps.”

The role of humans, says Dreckmeyr, is changing from actively controlling devices and machinery to passively monitoring them for potential failure. “Figure out what can be controlled in your environment and what can’t. Any decisions that relate to circumstances that cannot be controlled should be made by humans, not by machines,” Dreckmeyr says. “This is one reason why Gartner’s Maverick research has advised that CIOs who want the maximum benefit from smart machine solutions should ‘plan to deliver smart machine-enabled services that assist and are overseen by humans to achieve maximum benefit in the next three to five years, rather than those that are fully autonomous’,” he explains.

  1. Don’t assume a “Digital Twin” is the real thing 

Gartner defines a digital twin as “a dynamic software model of a physical thing or system that relies on sensor data to understand its state, respond to changes, improve operations and add value. Digital twins include a combination of metadata (for example, classification, composition and structure), condition or state (for example, location and temperature), event data (for example, time series), and analytics (for example, algorithms and rules).” Gartner posits that within three to five years, “hundreds of millions of things will be represented by digital twins”, and that organisations will “use digital twins to proactively repair and plan for equipment service, to plan manufacturing processes, to operate factories, to predict equipment failure or increase operational efficiency, and to perform enhanced product development”. Digital twins, Dreckmeyr says, are the result of the IoT, and should not be confused with the hardware-based sensors that provide the backbone. “You do actually still need to install those sensors,” he says. “People tend to forget the firstborn of the digital twin,” he only half-jokes.

If companies keep these simple principles in mind, they are unlikely to come unstuck with IoT implementations, Dreckmeyr says.

“It boils down to: think, plan, and only then execute in an area that can do with automation,” he ends.